Trace and Ultratrace Elemental Analysis - ACS Reagent Chemicals

Feb 28, 2017 - Flame AAS is a cost-effective analytical technique for the determination of common metallic impurities in most reagent chemicals, while...
1 downloads 9 Views 226KB Size
Procedure pubs.acs.org/doi/book/10.1021/acsreagents

Trace and Ultratrace Elemental Analysis Part 2, Analytical Procedures and General Directions eISBN: 9780841230460 Tom Tyner Chair, ACS Committee on Analytical Reagents James Francis Secretary, ACS Committee on Analytical Reagents

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

ABSTRACT The determination of individual elements can be performed by a number of atomic spectroscopic (AS) techniques, which differ mainly in their sensitivity, selectivity, sample throughput capability, and cost. The trace element techniques used in this book for measuring the purity of reagent chemicals include atomic absorption spectroscopy (AAS), inductively coupled plasma–optical emission spectroscopy (ICP–OES), and inductively coupled plasma–mass spectrometry (ICP–MS). Atomic absorption is still recommended for a group of alkali, alkaline earth, and transition elements, whereas plasma emission spectroscopy is now being phased in as the technique of choice for the determination of trace metal impurities in most of the reagent chemicals. It should also be emphasized that many laboratories are now using inductively coupled plasma–mass spectrometry for ultratrace metal determinations. Not only does this rapid multi-element technique have much lower detection limits than the other AS techniques, but also it offers the exciting capability of isotopic measurements.

GENERAL BACKGROUND The majority of routine trace metal analysis today is carried out by either atomic absorption, atomic emission, or mass spectroscopic techniques. Flame atomic absorption (flame AAS), electrothermal atomic absorption (ETAA), cold vapor atomic absorption spectroscopy (CVAAS), hydride generation atomic absorption spectroscopy (HGAAS), ICP–OES, and ICP–MS are the principal instrumental techniques used for trace metal analysis. In this book, AAS is recommended when the determination of a small number of elements in solution is required. Flame AAS is a cost-effective analytical technique for the determination of common metallic impurities in most reagent chemicals, while ETAA is used when the detection capability of flame AAS is not acceptable. CVAAS is used for the determination of mercury in ultratrace reagents, whereas HGAAS is used mainly for volatile elements such as selenium. ICP–OES is the technique of choice when rapid, multi-element analysis of solutions is required. A systematic changeover, starting with the determination of heavy metals in reagent chemicals, has begun with new plasma methods incorporated for over 50 reagents. The ultratrace reagents in this book are typically analyzed by ICP–OES for approximately 25 of the most laboratory-significant metallic impurities. Sample preparation typically involves dissolving and preconcentrating the chemicals by some kind of sample evaporation and/or matrix removal procedure before analysis. Alternatively ICP–MS can be used without the need for preconcentration techniques. The fundamental principles of atomic absorption, plasma optical emission, and plasma mass spectroscopy are well documented in the literature, so only the information pertaining to the specifications for the analyte elements of interest will be covered in this book. (See the bibliography on Trace Analysis beginning in [Part 6: List of Bibliographies; Physical Properties].) For that reason, specific instrumental parameters for each technique will not be given because operating procedures will vary depending on the design of the instrument.

© 2017 American Chemical Society

A

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

ACS Reagent Chemicals

Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

Reagents and Standards Precautions should be taken to avoid contamination. Sample containers and glassware should be carefully cleaned and stored. The highest purity water should be used, obtained preferably from a mixed-bed strong acid–strong base ionexchange cartridge capable of producing water with an electrical resistivity of 18 MΩ cm. It is strongly recommended that Teflon, fluorinated ethylene propylene (FEP), or ultraclean plasticware be used to avoid leaching of metals from laboratory glassware. Accurate trace element standard solutions should be used for the preparation of working calibration standards. These are commercially available from the manufacturers of high-purity standards and should be traceable to NIST standards. Concentrations of 1–10 g/L are usually recommended for storage. However, if lower concentration standards are being used, these solutions should be discarded after 24 h because of the strong likelihood of the metals being adsorbed by the walls of the glassware—unless experience has shown that they are stable for a longer period of time. Alternatively, appropriate plasticware should be used for long-term storage of these solutions.

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

The Ultratrace Environment The analysis of reagent chemicals intended for use in ultratrace applications must be done under “clean air” conditions because the limit of detection is very often dependent on the procedural “blank” obtained. The specifications used to determine the cleanliness of airborne particulates in a chemical laboratory are covered by ISO 14644, which replaced the older United States Federal Standard 209E. Conditions that conform to the air quality standards, as outlined in Class 1000 or better, where the airborne particle count is fewer than 1000 particles of 0.6 µm diameter per cubic centimeter of air, are generally considered sufficient for analysis of ultratrace reagents by flame AAS or ICP–OES. However, it is widely accepted that the more sensitive techniques are ETAA or ICP–MS. Class 100 air quality or better is need for part per trillion (ppt) detection limit. These ultraclean air conditions are usually achieved by the use of high-efficiency particulate filters in conjunction with proper laboratory design. Clean air conditions can be used in either an entire laboratory or in a laminar flow hood. Very often, in the absence of such facilities, clean air is supplied only to the specific area where contamination of a sample might occur. Extensive texts on these subjects are available, and the reader is strongly encouraged to examine these references before undertaking any ultratrace analysis.

Limit of Detection The determination of the limit of detection (LOD) for ultratrace analysis should be done using recognized methodologies, such as those used by the U.S. Environmental Protection Agency, SEMI International Standards, and American Public Health Association. Typical detection limits for the major AS techniques are shown in Table 2-9. Hg/Hydride refers to either mercury (Hg) or the hydride-forming elements including As, Bi Sb, Se, and Te. The Hg detection limit was determined by the cold vapor atomic absorption (CVAA) technique, whereas hydride generation atomic absorption (HGAA) was used to determine the detection limits of the volatile hydride-forming elements, using online chemical reduction techniques. Table 2-9. Atomic Spectroscopy Detection Limits (µg/L)

Element

Flame AA

Hg/Hydride

GFAA

ICP−OES

ICP−MS

Aluminum (Al)

45



0.1

1

0.0004*

Antimony (Sb)

45

0.15

0.05

2

0.0002

Arsenic (As)

150

0.03

0.05

1

0.0004

Barium (Ba)

15



0.35

0.03

0.00004

Beryllium (Be)

1.5



0.008

0.09

0.0003

Bismuth (Bi)

30

0.03

0.05

1

0.00002

Boron (B)

1000



20

1

0.001

Bromine (Br)









0.04

Cadmium (Cd)

0.8



0.002

0.1

0.00007

Calcium (Ca)

1.5



0.01

0.05

0.0003*

© 2017 American Chemical Society

B

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

ACS Reagent Chemicals

Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

Table 2-9. Atomic Spectroscopy Detection Limits (µg/L) (continued)

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

Element

Flame AA

Hg/Hydride

GFAA

ICP−OES

ICP−MS

Carbon (C)











Cerium (Ce)







1.5

0.00005

Cesium (Cs)

15







0.00005

Chlorine (Cl)









2

Chromium (Cr)

3



0.004

0.2

0.0003*

Cobalt (Co)

9



0.15

0.2

0.00006*

Copper (Cu)

1.5



0.014

0.4

0.0002*

Dysprosium (Dy)

50





0.5

0.0002

Erbium (Er)

60





0.5

0.0001

Europium (Eu)

30





0.2

0.00007

Fluorine (F)











Gallium (Ga)

75





1.5

0.00008

Gadolinium (Gd)

1800





0.9

0.0003

Germanium (Ge)

300





1

0.0006*

Gold (Au)

9



0.15

1

0.0001

Hafnium (Hf)

300





0.5

0.0003

Holmium (Ho)

60





0.4

0.00004

Indium (In)

30





1

0.00008

Iodine (I)









0.003

Iridium (Ir)

900



3.0

1

0.00009

Iron (Fe)

5



0.06

0.1

0.0005*

Lanthanum (La)

3000





0.4

0.00004

Lead (Pb)

15



0.05

1

0.00004*

Lithium (Li)

0.8



0.06

0.3

0.00005

Lutetium (Lu)

1000





0.1

0.00004

Magnesium (Mg)

0.15



0.004

0.04

0.0001

Manganese (Mn)

1.5



0.005

0.1

0.0001*

Mercury (Hg)

300

0.009

0.6

1

0.001

Molybdenum (Mo)

45



0.03

0.5

0.00008

Neodymium (Nd)

1500





2

0.0003

Nickel (Ni)

6



0.07

0.5

0.0002*

Niobium (Nb)

1500





1

0.00004

Osmium (Os)







6

0.00006

Palladium (Pd)

30



0.09

2

0.00003

Phosphorus (P)

75,000



130

4

0.04*

Platinum (Pt)

60



2.0

1

0.0001

Potassium (K)

3



0.005

1

0.001

Praseodymium (Pr)

7500





2

0.00003

Rhenium (Re)

750





0.5

0.0003

Rhodium (Rh)

6





5

0.00004

© 2017 American Chemical Society

C

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

ACS Reagent Chemicals

Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

Table 2-9. Atomic Spectroscopy Detection Limits (µg/L) (continued)

Element

Flame AA

Hg/Hydride

GFAA

ICP−OES

ICP−MS

Rubidium (Rb)

3



0.03

5

0.0002

Ruthenium (Ru)

100



1.0

1

0.0001

Samarium (Sm)

3000





2

0.0002

Scandium (Sc)

30





0.1

0.001

Selenium (Se)

100

0.03

0.05

2

0.0003*

Silicon (Si)

90



1.0

10

0.09

Silver (Ag)

1.5



0.005

0.6

0.00009

Sodium (Na)

0.3



0.005

0.5

0.0003

Strontium (Sr)

3



0.025

0.05

0.00007

Sulfur (S)







10

0.9*

Tantalum (Ta)

1500





1

0.00001

Terbium (Tb)

900





2

0.00003

Tellurium (Te)

30

0.03

0.1

2

0.0003*

Thorium (Th)







2

0.00005

Tin (Sn)

150



0.1

2

0.0002

Titanium (Ti)

75



0.35

0.4

0.0002*

Thallium (Tl)

15



0.1

2

0.00001

Thulium (Tm)

15





0.6

0.00003

Tungsten (W)

1500





1

0.00003

Uranium (U)

15,000





10

0.00002

Vanadium (V)

60



0.1

0.5

0.00007*

Ytterbium (Yb)

8





0.1

0.0001

Yttrium (Y)

75





0.2

0.00002

Zinc (Zn)

1.5



0.02

0.2

0.0007*

Zirconium (Zr)

450





0.5

0.00007

Note: “–” indicates that the technique is not optimal for that particular element.

• All detection limits are given in micrograms per liter and were determined using elemental standards in dilute aqueous solution. All detection limits are based on a 98% confidence level (3 standard deviations). • All atomic absorption detection limits were determined using instrumental parameters optimized for the individual element, including the use of System 2 electrodeless discharge lamps where available. • Data shown were determined on a PerkinElmer AA. • All Optima ICP–OES detection limits were obtained under simultaneous multi-element conditions with the axial view of a dual-view plasma using a cyclonic spray chamber and a concentric nebulizer. • Cold-vapor mercury detection limits were determined with a FIAS-100 or FIAS-400 flow injection system with amalgamation accessory. • The detection limit without an amalgamation accessory is 0.2 µg/L with a hollow cathode lamp and 0.05 µg/L with a System 2 electrodeless discharge lamp. (The Hg detection limit with the dedicated FIMS-100 or FIMS-400 mercury analyzers is ≤0.005 µg/L without an amalgamation accessory and ≤0.0002 µg/L with an amalgamation accessory.) Hydride detection limits shown were determined using a MHS-15 Mercury/Hydride system.

© 2017 American Chemical Society

D

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

ACS Reagent Chemicals

Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

• GFAA detection limits were determined on a PerkinElmer AA using 50 µL sample volumes, an integrated platform, and full STPF conditions. Graphite-furnace detection limits can be further enhanced by the use of replicate injections. • All ICP–MS measurements were performed on a NexION ICP–MS with a quartz sample introduction system using a 3 s integration time and 10 replicates in deionized water. Detection limits were measured under multi-element conditions in standard mode, except where denoted by an asterisk (∗). Detection limits denoted by * were performed in a Class 100 Clean Room using reaction mode with the most appropriate cell gas and conditions for that element in deionized water. Source: Reproduced with permission from PerkinElmer, 2013. Copyright 2013 PerkinElmer LAS.

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

ATOMIC ABSORPTION SPECTROSCOPY AAS is used when accurate determinations of only a few elements in solution are required. AAS has been chosen because it is a cost-effective analytical technique for trace metal analysis of many reagents contained in this book and is a technique capable of detection in the parts per million (ppm) range with flame atomization. However, if lower levels are encountered, then electrothermal atomization should be employed, which is capable of measurement at the parts per billion (ppb) level. Flame AAS is predominantly a single-element technique that uses a flame, such as from air–acetylene or nitrous oxide–acetylene, to generate ground-state atoms of the element of interest. The liquid sample is aspirated via a nebulizer and a spray chamber into the flame, where desolvation takes place and ground-state atoms are formed. The ground-state atoms in the sample absorb light of a specific wavelength from an element-specific, hollow cathode lamp (HCL) source. The amount of light absorbed, which is directly related to the number of ground-state atoms, is measured by a monochromator (optical system) and is detected by a photomultiplier tube or solid-state detector, which converts the photons into an electrical pulse. The sample absorbance signal for a particular element is compared to known calibration standards in order to determine the concentration of that element in the sample. Flame AAS typically uses about 2–5 mL/min of liquid sample. ETAA, also known as graphite furnace atomic absorption or GFAA, is also a single-element technique, although multi-element instrumentation is now available. It works on the same principle as flame AAS, except that the flame is replaced by a small heated tungsten filament or graphite tube. The other major difference is that a very small sample (typically 50 µL) is injected automatically onto the filament or into the tube rather than aspirated via a nebulizer and a spray chamber as in flame AAS. Because the ground-state atoms are concentrated in a smaller area than a flame, a higher degree of absorption takes place. The result is that ETAA offers about 100–1000 times lower detection limits than flame AAS. Although ETAA is widely used by the ultratrace element community, no methodology is presented in this book because flame AAS is sensitive enough to meet many of the reagent chemical specifications. However, if the ETAA technique is available, there is no reason why it cannot be used, assuming that the correct sample preparation and instrument methodology is applied. Conventional flame AAS spectrometers are equipped with a turret to accommodate multiple hollow cathode lamp sources. The burner is slotted and will vary in design and path length depending on the gases used. The monochromator will have a resolution of at least 0.1 nm. The signal from the photomultiplier tube detector is processed by a microprocessor and displayed on a digital readout or a terminal screen. To compensate for nonspecific absorption in the flame, a deuterium lamp background corrector or equivalent is used. In this book, typical instrument parameters, such as flame type, gas flows, absorption wavelengths, slit widths, and background correction methods, are given as guidelines in the procedures for individual reagents. However, the user should always refer to the manufacturer’s operator manual or application “cookbook” in order to achieve optimized conditions for a particular determination.

Sensitivity and Detection Limits Sensitivity in AAS is defined as the concentration of a test element in an aqueous solution that will produce absorption of 1% or 0.0044 absorbance units (AU). It is normally expressed in µg/mL or µg/g per 1% absorption. Note that this expression is related to the slope of the calibration curve. The sensitivity of the instrument is normally optimized during the method setup. A tabulation of expected absorbance vs standard concentration is shown in Table 2-10. There are many different definitions and procedures for determining detection limits depending on the data quality objectives of the analysis. The detection limit is often referred to as signal-to-background noise, and for a 99% confidence level, it is typically defined as three times the standard deviation of 10 replicates of the calibration blank (expressed in concentration units). To achieve

© 2017 American Chemical Society

E

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

ACS Reagent Chemicals

Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

the best detection limits, it is generally recognized that instrumental parameters will need to be optimized based on the application requirements.

© 2017 American Chemical Society

F

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

© 2017 American Chemical Society

589.0

460.7

213.9

Sodium

Strontium

Zinc

0.4

2.0

0.15

1.5

0.4

4.0

15.0

1.0

0.15

1.0

5.0

2.5

1.5

2.5

0.5

0.5

10.0

10.0

15.0

15.0

2.0

10.0

1.0

10.0

2.0

20.0

100.0

5.0

1.0

5.0

30.0

15.0

10.0

15.0

3.0

3.0

50.0

50.0

100.0

100.0

0.04

0.20

0.015

0.15

0.04

0.40

1.5

0.10

0.015

0.10

0.50

0.25

0.15

0.25

0.05

0.05

1.0

1.0

1.5

1.5

0.001

0.005

0.0004

0.0038

0.001

0.01

0.038

0.0025

0.0004

0.0025

0.013

0.0063

0.0038

0.0063

0.0013

0.0013

0.025

0.025

0.038

0.038

0.050

0.250

0.025

0.250

0.050

0.50

2.50

0.125

0.025

0.125

0.750

0.375

0.250

0.375

0.075

0.075

1.25

1.25

2.50

2.50

0.01–0.02

0.05–0.10

0.005–0.01

0.02–0.04

0.01–0.02

0.05–0.10

0.25–0.50

0.02–0.04

0.005–0.01

0.01–0.02

0.10–0.20

0.05–0.10

0.02–0.04

0.02–0.04

0.02–0.04

0.01–0.02

0.10–0.20

0.10–0.20

0.10–0.20

0.25–0.50

Linear Minimum (mg/ Minimum (mg/ Maximum (mg/ Recommended Range (mg/ L) 25 mL) 25 mL) Standard (mg/25 L) mL)

0.20–0.40

0.35–0.70

0.25–0.50

0.11–0.22

0.20–0.40

0.18–0.36

0.13–0.26

0.22–0.44

0.25–0.50

0.11–0.22

0.16–0.32

0.20–0.40

0.11–0.22

0.08–0.16

0.25–0.50

0.18–0.36

0.12–0.24

0.06–0.12

0.07–0.14

0.13–0.26

Expected AU

Notes: The data are based on instrument-specific conditions. In some instances, the recommended standard addition is lower than cited in this table. Dilution may be modified to meet maximum allowable limits. a Sensitivity check is the concentration giving approximately 0.2 AU. b Linear range is the upper concentration of the linear range. c Minimum is the concentration giving 0.02 AU (sensitivity check divided by 10). d Maximum is the upper limit in mg per 25 mL (linear range divided by 40).

766.5

328.1

Potassium

232.0

Nickel

Silver

313.3

285.2

Magnesium

Molybdenum

670.8

Lithium

279.5

217.0

Lead

Manganese

324.8

248.3

Copper

240.7

Cobalt

Iron

228.8

422.7

223.1

Bismuth

Cadmium

553.6

Barium

Calcium

396.2

217.6

Aluminum

Antimony

Wavelength (nm) Sensitivity Check (mg/L)

Element/ Analyte

Table 2-10. Expected Absorbance Values vs Standard Concentrations for a Group of Analytes

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

ACS Reagent Chemicals Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

G

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

ACS Reagent Chemicals

Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

Instrumentation The manufacturer’s instruction manual should be followed for detailed operating procedures and for matters of routine maintenance, cleaning, and calibration. The analyst should be familiar with the literature available in the public domain and be aware of the various interferences and potential sources of error. The procedures for the analysis of individual reagents in this book attempt to take into account the major interferences as much as possible. However, the analyst should always be aware of these interferences and ensure that the instrument being used is compensating for them.

Interferences The major interferences in atomic absorption spectrometry include the following:

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

• Sample transport: Surface tension, viscosity, acid content, and amount of dissolved solids may affect the nebulization rate and influence the sensitivity of an instrument by altering the concentration of the absorbing analyte in the flame. This is not such a serious problem in flame AAS, but if present, it can be compensated for by diluting the sample, matching matrix/acid content, or using the method of standard additions. • Chemical: With some samples, the analyte forms a strong chemical bond with other matrix components in the sample. This might necessitate the use of a much hotter nitrous oxide–acetylene flame or the addition of a complexing or releasing agent to change the chemistry of the analyte and/or the matrix. • Ionization: Some of the easily ionizable elements, such as sodium and potassium, readily ionize in a flame, which affects the number of ground-state atoms formed. This effect results in anomalous curvature of the calibration curve and can produce either positive or negative errors in the results. The addition of ionization buffers can minimize this effect. • Molecular absorption: Molecular absorption at the resonance wavelength or by light scattering is caused by very small particles in the flame. This type of interference becomes appreciable at 250–300 nm and increasingly severe at lower wavelengths. Scattering of radiant energy from particles in the flame is usually negligible above 300 nm or at sample matrix concentrations on the order of 10 g/L or less. Molecular absorption is usually compensated for by some kind of background correction technique. For an element with a resonance line in the ultraviolet region (≤350 nm), the correction can be performed with continuum source background correction using a deuterium arc lamp. For an element with a resonance line in the visible region (≥350 nm), an alternative approach known as the “adjacent line” technique needs to be used. In this method, the background absorption is determined at a wavelength adjacent to or close to the analyte resonance line, so the background absorption is measured at a wavelength where there is no significant atomic absorption by that element. Continuum source background correction is usually sufficient for flame AAS work, while an adjacent-line technique, such as Zeeman background correction, is preferred for ETAA work where high levels of structured background are sometimes encountered.

Analysis In this book, the predominant approach used for trace metal analysis of chemical reagents in solution by atomic absorption is the method of standard additions. This approach is used to ensure that the matrix has a minimal effect on the result. The method of standard additions uses the principle of making up the calibration standards in the same matrix as the sample by spiking the sample with fixed concentrations of analytes, corresponding to the number of points on the calibration curve. For standard additions to be applicable at low concentrations, it is essential that the sample solution, along with the spiked additions, generate absorbance signals in the linear part of the calibration curve, which is usually on the order of 0.01–0.50 AU. The method of standard additions does not eliminate the interference, but it does ensure that the elements of interest behave in a similar manner in both the sample and the calibration standards. However, if it is known that the chemical (matrix) effects are negligible, a calibration line using aqueous standards can be used when the slope and the slope of the standard addition calibration are identical. As indicated above, reagent-specific conditions for AAS analysis are listed under the individual reagents. This information includes, but is not limited to, wavelength, sample preparation, standard addition methodology, flame type, and background correction.

© 2017 American Chemical Society

H

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

ACS Reagent Chemicals

Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

Procedure for Flame AAS Prepare the sample for analysis by dissolving in water or by following a reagent-specific procedure. Prepare four solutions for standard addition calibration: reagent blank (which may be only water), sample solution, and two standard addition solutions. The recommended additions usually correspond to the specification limit and half the limit, although in some cases the limit and twice the limit are used. The signal for the reagent blank preparation can be subtracted from the signals of the three other solutions or the auto-zero feature of the instrument can be employed. The three results can be treated graphically or mathematically to determine the analyte concentration in the sample. Most modern instruments have standard addition software that will calculate this automatically. When several elements are to be determined, they can be added simultaneously to the standard addition solutions. A separate set of standard additions may be required if different sample weights are used in order to stay in the linear absorbance range.

Example.Figure 2-6 shows a typical procedure, as it appears in this book, that could be used for trace or ultratrace elemental analysis of a reagent chemical using the method of standard additions, which is illustrated here.

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

To a first set of three 25 mL volumetric flasks, add 2.0 mL (0.20 g) of sample stock solution. To two of the flasks, add the specified amounts of calcium and sodium shown in the table in Figure 2-6. Add 2 mL of 5% potassium chloride solution to all three flasks, and dilute to the mark with deionized water. To a second set of three 25 mL volumetric flasks, add 10.0 mL (1.0 g) of sample stock solution. To two of the flasks, add the specified amounts of potassium and strontium shown in the table in Figure 2-6. Dilute the contents of the three flasks to the mark with deionized water. Analyze the solutions by means of a suitable atomic absorption spectrophotometer, using the conditions outlined in the table in Figure 2-6. Calculate the trace metal content of the sample by the method of standard additions.

Figure 2-6. Sample specifications and tests for procedure using flame AAS

MERCURY DETERMINATION BY CVAAS The ultratrace determination of mercury in aqueous solution by atomic absorption is based on its reduction by a suitable reducing agent to the free element and transportation of the vaporized form of the element via a stream of inert gas into a quartz cell at room temperature (instead of the burner head used in flame AAS). A hollow cathode lamp can be used, but to maximize sensitivity, a more intense mercury electrodeless discharge lamp is used to generate wavelength-specific photons of light. The mercury ground-state atoms in the quartz cell then absorb the wavelength-specific light emitted by the electrodeless discharge lamp. Calibration is carried out in the normal way by comparing the absorbance of the sample with known mercury calibration standards. (The method of standard additions can also be used.) Commercially available mercury analyzers based on atomic fluorescence or amalgamation with a gold film may be substituted if the test method is validated. Care must be taken to guard against contamination of samples and solutions by mercury. Glassware should be washed with

© 2017 American Chemical Society

I

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

ACS Reagent Chemicals

Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

nitric acid (1:3), followed by thorough rinsing with water before analysis. However, dedicated plasticware or Teflon is strongly recommended for this type of analysis.

Procedure for Mercury Determination by CVAAS

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

Some details of sample preparation, including the use of reagent blanks, are described in the specific test for each reagent. Carry out all preparations in a fume hood. Samples are normally prepared in a set of three 100 mL volumetric flasks containing sample sizes described in the individual reagent monograph. Add mercury ion (Hg) to the second and third flasks. Use ultrapure reagents, including ultrapure water. If necessary, adjust sample weight and final volume according to the sensitivity of the mercury analyzer system. The manufacturer’s manual may suggest sensitivity or general guidelines. Transfer the sample solution, or an aliquot thereof, to the aeration vessel before adding any reducing agent. If the sample has been treated with permanganate, rinse the container with the specified quantity of hydroxylamine hydrochloride reagent solution, and add the rinsings to the aeration vessel with enough water to bring the level to the calibration line. Adjust the instrument to give a smooth baseline by flowing carrier gas through the bypass and cell and focusing the radiation of the 253.7 nm mercury resonance line through the cell onto the detector. Introduce stannous chloride solution as specified into the aeration vessel (generally 2.5% stannous chloride reagent solution, but follow the procedure recommended by the manufacturer’s manual), and set the valve to direct the carrier gas through the vessel and the cell. Allow the bubbling to continue until the instrument returns to its zero point. Before running the next sample, drain and rinse the aeration vessel. Periodically remove any deposit of stannic oxide.

SELENIUM DETERMINATION BY HGAAS HGAAS is commonly used for trace-level analysis of hydride-forming elements such as antimony, arsenic, and selenium because it gives better sensitivity for these elements than flame AAS. Selenium analysis, for example, using conventional flame AAS, is not sensitive because its analytical wavelength (196.0 nm) is too close to the vacuum ultraviolet region where absorption by air and flame gases occurs. In this book, HGAAS is used for the determination of selenium in ultratrace reagents and is based on selenium’s conversion to a volatile hydride (SeH2) by the sodium borohydride reagent and aspiration into an atomic absorption spectrometer fitted with a heated quartz atomization cell.

Equipment AAS systems equipped with a hydride reaction cell and an externally mounted, heated quartz cell or a quartz cell with an internal fuel-rich oxygen–hydrogen or air–hydrogen flame are recommended for this analysis.

Sensitivity and Interferences Quartz atomization cells are required for ultratrace analysis because they minimize background noise caused by the flame. The formation or presence of chlorine during hydride generation will result in low values due to the reoxidation of SeII to higher oxidation states.

Analysis Selenium determination is carried out using the method of standard additions. The reagent-specific conditions can be found under the individual reagent. Selenium is determined at the 196.0 nm wavelength and depends on the instrument setup. One microgram of selenium in a 100 mL volume gives approximately 0.12 AU, depending on the instrument used. Procedure for Selenium Determination by HGAAS Prepare the sample for analysis by following the reagent-specific procedure. The method of standard additions is the preferred analytical approach for selenium in order to minimize chemical- or matrix-induced interferences. However, for this method to be applicable at low concentrations, it is essential that the sample solution, along with the spiked additions, generate selenium absorbance signals in the linear part of the calibration curve. It should also be emphasized that the method of standard additions does not eliminate the interference, but it does ensure that the selenium behaves in a similar manner in both the sample and the calibration standards. However, if it is known that the chemical (matrix) effects are negligible, a calibration line using aqueous standards can be used instead of the standard addition calibration curve. The specific standard additions for selenium can be found under the individual reagents.

© 2017 American Chemical Society

J

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

ACS Reagent Chemicals

Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

INDUCTIVELY COUPLED PLASMA−OPTICAL EMISSION SPECTROSCOPY (ICP−OES)

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

Optical emission spectroscopy (OES) using inductively coupled plasma (ICP) as the excitation source is widely recognized as being an extremely rapid technique to carry out multi-element analysis in various sample matrices. Whereas atomic absorption (AA) is used to determine small numbers of elements in solution, ICP–OES is utilized for multi-element determinations or when high sample throughput is required. This technique is applicable to determine ultratrace elemental levels in many of the reagent chemicals specified in this book, including Ag, As, Bi, Cd, Cu, Hg, Mo, Pb, Sb, Sn, and all the elements that can be determined by AA (Ca, Ba, Na, K, Li, Mn, and Fe). The first phase of the changeover to the plasma emission technique, which has occurred in this edition of this book (Eleventh Edition), will lead to the eventual elimination of atomic absorption spectrometry. In order to achieve the specifications, many of the reagents can simply be diluted with deionized water, acidified with 5% nitric acid, made up to volume, and aspirated directly in the instrument’s sample introduction system. However, for analytical reagents that are more difficult to get into solution, an acid dissolution procedure using a hot plate or microwave digestion might be required. Additionally, for extremely low ultratrace specifications, some samples may need to be preconcentrated by evaporation under clean air conditions and made up to volume using 5% nitric acid prior to analysis by ICP−OES.

Fundamental Principles ICP–OES is a multi-element technique that uses inductively coupled plasma (ICP) to ionize and excite ground-state atoms to the point where they emit wavelength-specific photons of light characteristic of a particular element. The emitted photons are measured by a high-resolving optical system and photosensitive detector. This emission signal is directly related to the concentration of that element in the sample. The analytical temperature of ICP is about 6000–7000 K compared to a flame, which is typically 2500–4000 K. Detection limits for ICP–OES are generally very similar to flame atomic absorption (FAA) for the majority of elements but are much better for refractory and rare earth elements because of the higher excitation temperature. Sample volume requirements are on the order of 1 mL/min. The inductively coupled plasma is created in a quartz torch, which consists of three concentric tubes: outer tube, middle tube, and sample injector. The torch can either be one piece, where all three tubes are connected, or it can be a demountable design where the tubes and the sample injector are separate. Argon gas (known as plasma gas) is passed between the outer and middle tubes at a flow rate of about 12–17 L/min. A second argon gas flow (auxiliary gas) passes between the middle tube and the sample injector at about 1 L/min and is used to change the position of the base of the plasma relative to the tube and the injector. A third gas flow (nebulizer gas) also at about 1 L/min brings the sample, in the form of a fine droplet aerosol, from the sample introduction system (nebulizer and spray chamber) and physically punches a channel through the center of the plasma where the excitation takes place. The sample injector is often made from inert materials, like alumina, if highly corrosive materials are being analyzed. The principles of plasma formation are well-documented in the literature. First, a tangential (spiral) flow of argon gas is directed between the outer and middle tube of a quartz torch. A load coil (usually copper) surrounds the top end of the torch and is connected to a radio frequency (RF) generator. When RF power (typically 750–1500 W, depending on the sample) is applied to the load coil, an alternating current oscillates within the coil at a rate corresponding to the frequency of the generator, which in most commercial instruments is either 27 or 40 MHz. This RF oscillation of the current in the coil causes an intense electromagnetic field to be created in the area at the top of the torch. With argon gas flowing through the torch, a high voltage spark is applied to the gas causing some electrons to be stripped from their argon atoms. These electrons, which are caught up and accelerated in the magnetic field, then collide with other argon atoms, stripping off still more electrons. This collision-induced ionization of the argon continues in a chain reaction, breaking down the gas into argon atoms, argon ions, and electrons, forming the ICP discharge. The amount of energy required to generate argon ions in this process is on the order of 15.8 eV, which is enough energy to ionize and excite over 70 of the elements in the periodic table, which produce emission spectra mainly in the visible region (200–900 nm). However, spectrometers are now available that extend the wavelength coverage down into the low UV range (120–190 nm), where elements like S, P, Al, and the halogens can be determined. The emission produced from the plasma excitation process is then focused onto the entrance slit of a high-resolution optical system. Traditionally, these optics have either been based on a scanning monochromator (also known as a sequential system) or a dispersing polychromator (also known as a simultaneous system). The sequential system typically provides greater wavelength selection and flexibility, while the simultaneous system allows higher sample throughput. In both cases, very sensitive photomultiplier tubes are used as the detector to convert the photons into an electrical signal. More recently, spectrometers that use CCD (charge coupled device) or CID (charge injection device) solid-state detectors have become commercially available. These instruments are typically equipped with high-resolution Echelle gratings that disperse the

© 2017 American Chemical Society

K

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

ACS Reagent Chemicals

Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

spectrum in two dimensions, although technology is also available that uses traditional Rowland Circle technology with multiple solid-state detectors. The benefit of coupling solid-state detection to high-resolution two-dimensional spectrometers is that virtually any wavelength can be accessed very quickly, which makes them very flexible and extremely fast. In addition to the different optical designs, it should also be noted that there are two main types of ICP-viewing configurations available—a conventional radial design, where the torch is positioned vertically (sometimes known as side-on viewing), and an axial configuration, where the torch is positioned horizontally (sometimes known as end-on viewing). As mentioned previously, radial-view ICP can typically achieve similar detection limits to flame AA for the majority of elements, with the exception of the refractory and rare earth elements. The benefit of the axial design is that more photons are seen by the detector, and as a result, detection limits can be as much 5–10 fold lower, depending on the design of the instrument. However, the disadvantage of this approach is that more severe matrix interferences are observed, which means that matrix matching is very critical. To compensate for this limitation in the end-on plasma design, manufacturers now offer dual-view instruments.

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

Detection Limits, Sensitivity, and Linear Dynamic Range Limits of detection, sensitivity, and linear dynamic range should be investigated and established for each individual analyte line or wavelength on the particular instrument being used for the analysis of the individual reagent. Refer to the ICP–OES “typical” detection limits shown in Table 2-9 to ascertain whether the technique is suitable for the element being determined at the specification defined in the particular reagent chemical being tested. The required detection capability will also depend on the sample preparation technique being used. For example, the weight of the sample and the final volume will impact the concentration of the heavy metals (by ICP–OES) in solution. Refer to Table 2-11 for more information. Consideration should also be given to the linear dynamic range of the particular emission lines being selected for the analysis. Although this will vary slightly depending on the type of instrument and viewing configuration used, most ICP–OES systems on the market are capable of achieving linear dynamic ranges of approximately 5 orders of magnitude. So if the analyte detection limit is 1 ppb, the calibration line will be linear up to 100 ppm. Also refer to the manufacturer’s application-specific methodology together with reference literature in the public domain for sensitivity, detection limits, and linear dynamic range of the emission lines being used for the analysis. Table 2-11. Elemental Concentrations (ppm) in Solution Based on Specification and Sample Preparation (weight/ volume) Used To Get the Sample into Solution

Sample Weight/Final Volume Specification of Heavy Metals (by ICP−OES) [ppm (%)]

0.5 g/100 mL

1.0 g/100 mL

2.0 g/100 mL

5.0 g/100 mL

Table 2-11. Elemental Concentrations (ppm) in Solution Based on Specification and Sample Preparation (weight/ volume) Used To Get the Sample into Solution

Elemental Concentration in Solution (ppm) 1 ppm (0.0001%)

0.005

0.01

0.02

0.05

5 ppm (0.0005%)

0.025

0.05

0.10

0.25

10 ppm (0.001%)

0.050

0.10

0.20

0.50

100 ppm (0.01%)

0.50

1.0

2.0

5.0

1000 ppm (0.10%)

5.00

10.0

20.0

50.0

Interferences In general, ICP–OES is mostly prone to spectral-type overlaps and matrix suppression effects produced by other matrix components in the sample. These types of interferences must be fully investigated in the method-development stage

© 2017 American Chemical Society

L

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

ACS Reagent Chemicals

Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

before accurate multi-element determinations can be carried out. Special attention must also be paid to the possibility of external contamination during sample preparation and analysis, especially if ultratrace determinations are being carried out. These external contamination sources cannot only affect the accuracy of analyte recoveries but also impact the detection capability by generating unforeseen spectral and matrix-induced interferences. The use of an internal standard (IS) is required, as it compensates for the physical and matrix interferences resulting from differences in sample transport efficiencies and/or excitation conditions between the sample matrix and calibration solutions. The choice of internal standard and wavelength should be left to the analyst, as the optimum selection will depend on a number of different factors. However, it is very important that the internal standard element chosen is not present in the samples. It is also desirable that the excitation potential of the internal standard is similar to that of the analytes. Three of the most widely used internal standards used in ICP emission include scandium (Sc), yttrium (Y), and lanthanum (La). Table 2-12 shows some suggested wavelengths in descending order of sensitivity.

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

Table 2-12. Some Suggested Internal Standards and Wavelengths To Use To Compensate for Matrix-Induced Interferences

Internal Standard

Wavelength (1) (nm)

Wavelength (2) (nm)

Wavelength (3) (nm)

Scandium (Sc)

361.38

357.25

424.68

Yttrium (Y)

371.02

324.22

361.10

Lanthanum (La)

408.67

379.48

407.74

If an unknown matrix is encountered, the method of standard additions should always be considered as an option to compensate for large matrix differences between samples and standards. However, it is important to emphasize that unlike AA standard additions, where only one analyte is spiked into each solution, in ICP–OES standard additions, all the analyte elements being determined need to be spiked into the solutions. This makes the analysis very time consuming, so this approach should only be used if all other matrix compensation approaches have failed to work. Shifts in background intensity levels from recombination effects and/or molecular band spectral contributions may be corrected by the use of an appropriate background correction technique. Direct spectral overlaps are best addressed by selecting alternative wavelengths. Spectral interference studies should be conducted on all new matrices to determine whether interference correction factors should be applied to concentrations obtained from certain spectral line intensities to minimize biases. Some instrument manufacturers offer software programs that mathematically correct for direct spectral overlaps and background shifts. It is therefore highly recommended that the spectral areas around the analyte emission lines be examined to ascertain whether background correction points are needed or even if alternative lines should be selected. If a fixed channel, polychromator-type spectrometer using photomultiplier tube detection is being used for the analysis, it is recommended that inter-element correction factors be applied, based on the expected interfering elements present in the samples. The user may also choose multiple wavelengths if the spectrometer has the capability to help verify that emission line selection is optimized for the particular reagent chemical being analyzed.

Instrumentation An ICP–OES instrument with sufficient resolution to separate the relevant analyte emission lines of interest is required for this analysis. No specific procedure will be given because operating conditions will vary depending on instrument design. However, it is important that the optimum RF power, plasma/nebulizer gas flows, and sample introduction parameters are used for samples being analyzed. Also, ensure that the instrument has reached temperature equilibrium before analysis begins. To ensure that the instrument is achieving adequate sensitivity, the user should check its performance specifications on a routine basis. If in doubt, refer to the instrument operator’s manual for guidance. The instrument should have the ability to monitor one or more of the wavelengths for each analyte shown in Table 2-13, in addition to at least one of the internal standard wavelengths shown in Table 2-12. Table 2-13. List of Recommended Analyte Wavelengths in Descending Order of Sensitivity

Element

Wavelength (1) (nm)

Wavelength (2) (nm)

Wavelength (3) (nm)

Aluminum (Al)

396.15

308.22

394.40

Antimony (Sb)

206.83

217.58

231.15

Arsenic (As)

188.98

193.70

234.98

© 2017 American Chemical Society

M

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

ACS Reagent Chemicals

Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

Table 2-13. List of Recommended Analyte Wavelengths in Descending Order of Sensitivity (continued)

Element

Wavelength (1) (nm)

Wavelength (2) (nm)

Wavelength (3) (nm)

Bismuth (Bi)

223.06

222.82

306.77

Barium (Ba)

455.40

233.53

493.41

Boron (B)

249.77

249.67

208.89

Cadmium (Cd)

214.44

226.50

228.80

Calcium (Ca)

393.37

396.85

315.88

Chromium (Cr)

205.55

267.72

283.56

Cobalt (Co)

228.63

238.89

230.79

Copper (Cu)

324.75

327.39

213.60

Iron (Fe)

238.20

259.94

234.35

Lead (Pb)

220.35

283.31

182.14

Lithium (Li)

670.78

610.37

460.29

Magnesium (Mg)

279.55

279.07

285.21

Manganese (Mn)

257.60

259.37

294.92

Mercury (Hg)

184.89

194.16

253.65

Nickel (Ni)

231.60

221.65

232.00

Molybdenum (Mo)

202.03

204.60

284.82

Potassium (K)

766.49

769.90

404.72

Silicon (Si)

212.41

251.61

288.16

Silver (Ag)

328.07

338.29

243.78

Sodium (Na)

589.59

588.98

330.24

Strontium (Sr)

407.77

421.53

460.73

Tin (Sn)

189.99

235.49

284.00

Titanium (Ti)

334.94

336.12

337.28

Vanadium (V)

292.40

290.88

310.23

Zinc (Zn)

213.86

206.20

202.55

Zirconium (Zr)

343.82

339.20

257.14

Calibration Multi-element standard solutions should be prepared on the day of analysis by dilution of a suitable stock standard solution. They should be matrix-matched to the sample solution, which is typically 5% nitric acid (HNO3). Note: It is important that all stock solutions be traceable to an internationally recognized standards organization such as NIST (National Institute of Science and Technology). The elemental components should be selected based on the analytical requirements and the chemistry of the individual elements to provide stable solutions for the required concentration ranges. Analyze the samples using matrix-matched calibration standards and a blank. The number and range of calibration standards should be selected based on the reagent specification and the sample preparation dilution factor. Two standards should be used: one at the lower end and one at the higher end of the calibration range. It is fairly typical to use 0.50 and 2.00 ppm. Refer to Table 2-11 for some additional guidance.

© 2017 American Chemical Society

N

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

ACS Reagent Chemicals

Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

Procedure The specifications of many of the reagent chemicals can be achieved by a straight forward approach, where a known amount of sample is dissolved in deionized water, acidified with 5% HNO3, and analyzed directly by ICP–OES using internal standardization (IS). However, the achievable detection capability can be improved by preconcentration of the sample using evaporation procedures under clean air conditions. Reliable and accurate results can also be assured by the use of a quality control (QC) sample or certified reference material (CRM). The sample size will depend on the detection limits required and the capability of the instrument used. A spectral background correction technique like inter-element correction (IEC) or multicomponent spectral subtraction (MSS) might be required, especially with spectrally complex samples that contain high levels of transition metals such as iron (Fe), copper (Cu), chromium (Cr), nickel (Ni), cobalt (Co), molybdenum (Mo), and tungsten (W). Whatever method is used, before quantitative analysis can begin, rigorous method development has to be carried out to ensure that the selected wavelengths and instrument sampling conditions are appropriate for the reagent chemical being analyzed.

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

Determination of Heavy Metals by ICP−OES It is recommended that the ICP used in the analysis be standardized using a matrix-matched blank and two standards in the range from 0.5 to 2 ppm each of silver, arsenic, bismuth, cadmium, copper, mercury, molybdenum, lead, antimony, and tin. An internal standard should also be run as part of the analysis. Scandium or yttrium is recommended, but any appropriate internal standard may be used.

Sample Preparation Dissolve the quantity of sample suggested in the individual monograph in approximately 50 mL of water in a 100 mL volumetric flask. Acidify the sample with 5 mL of high purity nitric acid, add 0.1 mL of 1000 ppm gold standard solution, and dilute to 100 mL with water. Divide into two aliquots of 50 mL each. Analyze one aliquot by ICP–OES for Ag, As, Bi, Cd, Cu, Hg, Mo, Pb, Sb, and Sn. Spike the second aliquot with 0.1 mL of 100 ppm of a suitable multi-elemental standard and analyze. Calculate the spike recovery based on the results from solution with and without the spike and a 5% nitric acid blank, and report the results obtained for each analyte in the unspiked aliquot. Note: It is well recognized that the addition of gold is used for stabilization purposes in the determination of mercury. However, it is strongly recommended that samples and standards be analyzed immediately after preparation. For more information, refer to the following U.S. Environmental Protection Agency document, “Mercury Preservation Techniques” (http://www.epa.gov/esd/factsheets/mpt.pdf, accessed April 18, 2013).

Spike Recovery Calculation—Method Validation Concentration of the target element (Ct) can be calculated as follows:

Ct = Cs × (It – Ib)/(Is – It), where Cs = concentration of the spiked solution, It = intensity of target solution, Is = intensity of spiked solution, and Ib = intensity of blank. Spike recovery can be calculated as follows: Spike recovery = (Cs – Ct)/Cs, where Ck = known concentration of spike. Acceptable spike recoveries are 75–125%. Once all acceptable spike recoveries have been achieved, carry out five replicates of the sample and five replicates of the blank. The standard deviation (%SD) of the blank readings should be multiplied by 3 to give the method detection limit. All determinations less than this value should be reported as below the detection limit. If more than one wavelength has been chosen for quantitation, report the one that gives the lowest %RSD.

Calculation of Heavy Metals (by ICP−OES) Results The reported results should be the summation of all concentrations of targeted elements. This result may not exceed the specification given for heavy metals (by ICP–OES) in each monograph.

© 2017 American Chemical Society

O

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

ACS Reagent Chemicals

Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

Preconcentration of Ultratrace-Grade Reagents for Analysis by ICP−OES In a clean air environment, place 100 g of sample in one 100 mL PTFE (or equivalent) evaporating dish and a reagent blank in another. Slowly evaporate on a hot plate, avoiding loss of sample by effervescence or spattering, until approximately one drop of liquid remains. Do not allow the sample to evaporate to complete dryness. Cool. Transfer to a 100 mL volumetric flask. Acidify the sample with 5 mL of high purity nitric acid, add 0.1 mL of 1000 ppm gold standard solution, and dilute to 100 mL with water. Analyze the reagent chemical by ICP–OES using suitable calibration standards and a reagent blank of the specified elements listed in the individual monograph of the ultratrace reagent. Refer to Table 2-13 for recommended analyte wavelengths, together with Table 2-12 for a suitable internal standard. Note: It is strongly recommended that samples and standards be analyzed immediately after preparation for the stability reasons mentioned earlier.

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

ICP−MS Since it was first developed more than 30 years ago, ICP mass spectrometry (ICP–MS) has enabled trace metal detection capability at the parts per trillion (ppt) level. Even though ICP–MS can broadly determine the same suite of elements as the other atomic spectroscopic techniques, it has clear advantages in its multi-element characteristics, speed of analysis, detection limits, and isotopic capability. It is acceptable to use this technique for the ultratrace reagents for many of the reagent chemicals as long as method validation has been done. Several different ICP–MS designs are available today. They share many components, such as the nebulizer, spray chamber, plasma torch, and detector, but can differ quite significantly in the design of the interface, ion focusing system, mass separation device, and vacuum chamber. As in ICP–OES, the sample is delivered into a nebulizer, where it is converted into a fine aerosol with argon gas. The fine droplets of the aerosol are separated from larger droplets by means of a spray chamber. The fine aerosol then emerges from the spray chamber and is transported into the plasma torch via a sample injector. The plasma torch in ICP–MS functions differently from in ICP–OES. The plasma is formed in a similar way, but this is where the similarity ends. In ICP–OES, the plasma is used to generate photons of light, whereas in ICP–MS, the plasma torch is used to generate positively charged ions. The production and detection of large quantities of positively charged ions gives ICP–MS its characteristic low parts per trillion detection capability—about three or four orders of magnitude better than ICP–OES. Once the ions are produced in the plasma, they are directed into the mass spectrometer via the interface region, which is maintained at a vacuum of 1–2 Torr with a mechanical pump. This interface region consists of two metallic cones called the sampler and the skimmer cones, each with a small orifice (0.6–1.2 mm) to allow the ions to pass through to the ion optics, where they are guided into the mass separation device. After the ions have been successfully extracted from the interface region, they are directed into the main vacuum chamber by a series of electrostatic lenses, called ion optics. The operating vacuum in this region is maintained at about 10–3 Torr with a turbomolecular pump. There are many different designs of the ion optic region, but they serve the same function, which is to electrostatically focus the ion beam toward the mass separation device while stopping photons, particulates, and neutral species from reaching the detector. The ion beam containing all the analyte and matrix ions exits the ion optics and passes into the mass separation device, which is maintained at a vacuum of 10–6 Torr with a second turbomolecular pump. The many mass separation devices include quadrupole, magnetic sector, time-of-flight, and collision/reaction cell technology. Their principles of operation differ, but they all allow analyte ions of a particular mass-to-charge ratio to pass through the detector and reject all the nonanalyte ions, interfering species, and matrix components. The last step is converting the ions into an electrical signal with an ion detector. The most common design today is called a discrete dynode detector, which contains a series of metal dynodes along the length of the detector. In this design, when the ions emerge from the mass filter, they impinge on the first dynode and are converted into electrons. As the electrons are attracted to the next dynode, electron multiplication takes place, which results in a very high stream of electrons emerging from the final dynode. This electronic signal is then processed by the data-handling system and converted into analyte concentration using ICP–MS calibration standards. Most detection systems used can handle up to

© 2017 American Chemical Society

P

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2

ACS Reagent Chemicals

Procedure

pubs.acs.org/doi/book/10.1021/acsreagents

eight orders of dynamic range, which means they can be used to analyze samples from the parts per trillion level up to a few hundred parts per million.

Interferences

Downloaded by CORNELL UNIV on May 6, 2017 | http://pubs.acs.org Publication Date (Web): February 28, 2017 | doi: 10.1021/acsreagents.2006

Although the detection capability of ICP–MS is significantly better than that of other AS techniques, it is prone to certain interferences. Although these interferences are reasonably well understood, it can often be difficult and time consuming to compensate for them, particularly in complex sample matrices. Having prior knowledge of the interferences associated with a particular set of samples will often dictate the sample preparation steps and the instrumental methodology used to analyze them. Interferences are generally classified into two major groups: spectral-based and matrix-based. Each has the potential to be problematic, but modern instrumentation and good software, combined with optimized analytical methodologies, have minimized their negative impact on trace element determinations by ICP–MS. -

Spectral interferences. These are probably the most serious type of interferences in ICP–MS. The most common ones are known as polyatomic or molecular spectral interferences, which are produced by the combination of two or more atomic ions. They are caused by a variety of factors but are usually associated with argon gas, matrix components in the solvent or sample, other elements in the sample, or entrained oxygen or nitrogen from the surrounding air. Simple polyatomic interferences are handled using mathematical correction equations, whereas severe ones have to be compensated for by using either cool plasma conditions, collision/reaction cells, or in extreme cases by using a highresolution spectrometer.

-

Matrix interferences. These are classified into three different categories. The first and simplest to overcome is a sample transport effect and is a physical suppression of the analyte signal brought on by the level of dissolved solids or acid concentration in the sample. The second type of matrix suppression is caused when the sample affects the ionization conditions in the plasma discharge. This results in the signal being suppressed by varying amounts, depending on the concentration of the matrix components. The third type of matrix interference, called a space-charge effect, is caused by suppression of a high-mass matrix element on the transmission of a lower mass analyte ion through the ion optics. The susceptibility of ICP–MS to matrix-induced interferences means that internal standardization is nearly always used for the quantitation of unknown samples.

Instrumentation and Methodology An instrument with sufficient resolving power to separate the mass spectrum of the elements of interest is required. No specific methodology is provided here because operating procedure will vary depending on whether the instrument design is based on quadrupole, collision/reaction cell, double-focusing magnetic sector, or time-of-flight technology. The manufacturer’s instruction manual should be followed for detailed operating procedures and for matters of routine maintenance, cleaning, and calibration. The analyst should be familiar with current literature and aware of the various interferences and potential sources of error for the analytes of interest.

© 2017 American Chemical Society

Q

ACS Reagent Chemicals ACS Reagent Chemicals; American Chemical Society: Washington, DC, 2017.

DOI:10.1021/acsreagents.2006 ACS Reagent Chemicals, Part 2